摘要
非侵入式负荷监测是自动化设备用电安全监测的重要技术。针对常用电器存在功率值接近导致负荷分解准确率低的问题,原有基于低频数据的方法无法对其有效分解,本文提出了一种基于SV-ISSA(State vector-Improved Sparrow Search Algorithm)的非侵入式负荷分解方法。基于事件检测后得到的功率波形,使用DTW算法提取电器的典型功率矩阵,利用改进层次聚类算法将功率矩阵转换为状态向量,并利用滑动窗口提取电器超状态,对超状态进行缩减并获取其对应功率矩阵。此时负荷分解问题可转换为求解最优组合问题,然后通过改进的麻雀搜索算法求解各个电器超状态的运行状况,从而得到最终分解结果。实验结果表明,采用所提方法可有效提高负荷分解准确率,且能够准确处理具有相似功率范围同时运行的多种电器。
Non-intrusive load monitoring is an important technology for power safety monitoring of automation equipment. Given the low accuracy of load decomposition caused by the close power value of household appliances, the original method based on low-frequency data cannot decompose them effectively. This paper proposes a non-invasive load decomposition method based on SV-ISSA(state vector improved sparrow search algorithm). Based on the power waveform obtained after event detection, the DTW algorithm is used to extract the typical power matrix of electrical appliances, the improved hierarchical clustering algorithm is used to convert the power matrix into a state vector, and the sliding window is used to extract the super state of electrical appliances, reduce the super state and obtain its corresponding power matrix. At this time, the load decomposition problem can be transformed into solving the optimal combination problem, and then the operation status of each electrical appliance in the super state can be solved by the improved sparrow search algorithm, to obtain the final decomposition result. The experimental results show that the proposed method can effectively improve the accuracy of load decomposition, and can accurately deal with a variety of electrical appliances operating at the same time with similar power ranges.
出处
《自动化博览》
2022年第11期72-77,共6页
Automation Panorama1
基金
杭州电力设备制造有限公司科技项目(YF211601)。
